[EM] MIT News: Math of elections says voters win with 'winner take all'

[raphfrk at netscape.net]

  from the article:
 
 > Natapoff would count popular votes cast for any candidate 
  > vote-for-vote for the state's winner: If Florida casts 6 
  > million votes for all the candidates, its winner should receive > precisely 6 million electoral votes plus the popular-vote equivalent 
  > of two senatorial electoral votes--a quarter of the popular vote 
  > in the average state, or about half a million votes now. 
 > "This system would empower voters in poorly contested states, 
 > who could withhold their vote from the state's winner by casting 
 > a blank ballot," Natapoff wrote. "The dominant candidate would 
 > need (acceptance from his opposition) or risk losing 40 percent 
 > of the state's electoral votes." It would give 80 million impotent 
 > voters in those states an immediate impact on presidential elections. 
 > It is the only basic change we need or dare make, he says. 
   
 My interpretation is that in a 2 candidate race, the voter can 
 vote 3 ways.
 
 - candidate A (A-type)
 - candidate B (B-type)
 - blank (X-type)
 
 His plan is that the winning candidate gets
 (A+B) votes. The X votes are just ignored.
 
 The X option has the same effect as not voting, so it is reasonable
 to assume that voters who want to vote that way just won't vote.
 
 It seems to me that this system has plurality like disadvantages. 
 However, it is even 'better' than plurality, in that it results in
 1 party domination of each State rather than 2 party domination.
 
 The voter's best strategy is
 
 1) estimate who will win your State (SW)
 
 2) estimate the top 2 in the national election (NW) and (2nd)
 
 3) vote:
 
 if SW is your favorite of the top 2, vote SW
 if SW is your least favorite of the top 2, vote blank (or don't vote)
 if SW is another candidate, vote SW if SW is better than expectation of election
 Otherwise, vote blank (or don't vote)
 
 The top 2 clauses are the most likely. If everyone follows the
 strategy, all the SW's will be from top 2.
 
 The effect is identical to the NPV. If you are a Democrat
 in a Republican State, you will just not vote. If you are a
 Republican in a Democrat State, you will just not vote. It
 is likely that the State winners would be getting 90%+ of 
 the votes in their State.
 
 The likely effect is to keep States locked for one party or 
 the other. If you haven't voted in the last 10 elections,
 would you bother to vote in the current one where your
 party has finally obtained a plurality. 
 
 Likely, alot of people wouldn't so, even a candidate with 
 60% support in polls in a State wouldn't be the most 
 likely to win. The candidate from the 'main' party in 
 the State would have the best chance.
 
 It possibly would come down to how much use people 
 make of the blank ballot option. If in the last election 
 a good chunk of the weaker party's supporters voted blank, 
 then they have a better chance of being convincing 
 about who can win.
 
 However, the likely effect is that the weaker party
 in each State would just die like 3rd parties do.
 
 The attitude of the weaker party's supports would probably
 progress from "I'll vote so that the State does well 
 and anyway, the votes go to the State winner" to 
 "my party never has a chance because none of its
 supporters vote. If I vote I am just giving votes to
 the other party"
 
 The overall effect is NPV, but with only supporters
 of the dominant party in each State being allowed to
 vote. It may not be an entire disaster. The other
 party can still campaign there to try to get people
 to not vote. This means that it is a NPV system, it
 is just that the threshold for winning is not at 50%.
 This means that it isn't very democratic.
 
 For example, consider a 2 State country. One State
 has 10 voters and the other 6.
 
 Party A controls the 10 State and Party B controls
 the 6 State. All voters will vote if they support the
 major party in their State and not otherwise.
 
 Party A will get
 10: Pa10
 6: 0
 
 Party B will get
 10: 0
 6: Pb6 = 6 - Pa6
 
 Determining who will win
 
 (Party A - Party B) is
 
 Pa10 - (6 - Pa6)
 
 =
 
 (Pa10 + Pa6) - 6
 
 =
 
 (Pa_Total) - 6
 
 If this is positive, party A will win. If party A has 6 votes
 nationally and party B has 10 votes nationally, it will be a tie.
 This means that party A can win with only 37.5% of the vote.
 It doesn't matter how they are distributed nationally. However, 
 they would need to maintain reasonable support or risk a State 
 switching. (which would be alot less likely than now).
 
 > Small states cancel each other in a close election. 
 > The greater coherence of large states under winner-take-all, 
 > Natapoff claims, gives them much greater national power per 
 > vote--in proportion to the square root of their size--than the 
 > same number of electoral votes in small states. That, he 
 > believes, is why senatorial electoral votes have worked for 
 > two centuries and are still needed.
 
 The irony is that since it is not actually winner takes all, 
 including the Senate votes is unfair to the larger States.
 All it does is shift the percentage of the national popular
 vote needed for the 2 parties to win.
 
 
  Raphfrk
 --------------------
 Interesting site
 "what if anyone could modify the laws"
 
 www.wikocracy.com 
________________________________________________________________________
Check Out the new free AIM(R) Mail -- 2 GB of storage and industry-leading spam and email virus protection.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20070414/d74e6620/attachment-0001.htm>